Problematic surfaces extracted using watershed

[sdeslauriers]

For some subjects of the HCP, the 3 layer bem model raises an error in mne when trying to extract surfaces.

For example, running make_bem_model on the output of watershed_bem on subject 109123 of the HCP raises the following error:

RuntimeError: Surface inner skull is not completely inside surface outer skull

Checking the surfaces using mne freeview_bem_surfaces shows that the inner skull intersects the outer skull and skin surfaces. In addition, the inner skull surface has loops. This also happens with other subjects.

Using the --atlas option of mne watershed_bem helps with the loops, but still produces intersecting surfaces.

For now, the fix is to use a 1 layer bem by setting the conductivity option of mne.make_bem_model to a list with 1 value (e.g. [0.3]).

This issue is meant to document this problem, as true fixes are not obvious.

See also: https://github.com/mne-tools/mne-python/issues/5565

[dengemann]

Thank you! I think there is little to do as this is what we get from the HCP. Perhaps we could do a page somewhere were we document such issues. On Mon 1 Oct 2018 at 17:14, Samuel Deslauriers-Gauthier < notifications@github.com> wrote:

For some subjects of the HCP, the 3 layer bem model raises an error in mne when trying to extract surfaces.

For example, running make_bem_model on the output of watershed_bem on subject 109123 of the HCP raises the following error:

RuntimeError: Surface inner skull is not completely inside surface outer skull

Checking the surfaces using mne freeview_bem_surfaces shows that the inner skull intersects the outer skull and skin surfaces. In addition, the inner skull surface has loops. This also happens with other subjects.

Using the --atlas option of mne watershed_bem helps with the loops, but still produces intersecting surfaces.

For now, the fix is to use a 1 layer bem by setting the conductivity option of mne.make_bem_model to a list with 1 value (e.g. [0.3]).

This issue is meant to document this problem, as true fixes are not obvious.

See also: mne-tools/mne-python#5565 https://github.com/mne-tools/mne-python/issues/5565

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/mne-tools/mne-hcp/issues/60, or mute the thread https://github.com/notifications/unsubscribe-auth/AB0fit27ZMaJcnmbneDaUHtYxPaU7AHmks5ugjFFgaJpZM4XCSXq .